TWI270059B - Data player - Google Patents

Abstract

The present invention relates to a data player, which sequentially processes a reproduced signal of a reproducing head with a simulated filter, an A/D converter and an interpolator and provides the processed signal to a digital balancer. After the digital balancer digitally balances the inputted digital signal, the balanced signal is outputted to a next-stage Viterbi decoder. The digital balancer includes a variable BPF, a variable FIR filter, and the characteristics of the filter can be adjusted. In the player that reproduces digital data, a balancing processing of the reproduced digital signal can be implemented in a versatile manner.

Description

12/UU3y IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to equalization processing of data. , 寸 疋 疋 资料 资料 【 【 【 【 【 【 【 【 【 【 【 ♦ ♦ ♦ ♦ ♦ ♦ 现在 现在 有 有 有 有 有 有 有 有 有 ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ In the case, the error is reduced by compensating for one-line compensation in the deterioration of the signal or the tape type of the Tezizaki regeneration system. ', the difference in characteristics of the singer, etc. The fourteenth figure shows the existing block diagram with the equalizer. The reproduction circuit 1〇3 is constituted by a clothes spoon and outputs an image signal and a sound signal to the track of the band 1 (1). The visitor cried ς ς ς 十 5 5 5 彳 彳 彳 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心 心Through the wave of a (aUpass); control the high-frequency group delay of the full-pass wave frequency lifting filter ^ material from the amplifier (9) 贱 high frequency = money to compensate the output of the long-band group delay control (four) all-pass chopping. The low-frequency group delay·_ all-pass money device, for the double-grain of the tape, Fu Zhensheng performs the reconnaissance on the low-frequency group delay of the regenerative nickname, and the sub-output to the high-frequency group delay control In the pass filter. High-frequency group All-pass m for delay control controls the high-frequency group delay of the analog reproduction signal, and then performs integral processing and outputs it to A/D1〇9. 20 127〇〇59 A/D 109 ‘will come from. The signal is output to m (there is ==;^ is converted to a number than the regenerative signal. A/foot, and from PLL113^) and 111 and PLL113 5 10 15 20 are sampled. ...the ten-receiving signal FIR filter, the waver 111, the pre-coded decoder 115 of the star PR4, the estimation circuit m and the Viterbi decoder 121=. = weekly processing. The characteristics of the output to the scale 107 are evaluated, and the surname is: path 117, for the uniform circuit U9, and the characteristics of the equalizer 1〇7 are adjusted. : Out to the equalizer controller (2), using the Viterbi algorithm to extract the stone from pR4, the Viterbi decoding sample 1 bit of the digital signal is detected, and the second = output command will be the i-night signal processing circuit 123' according to the Viterbi 丨^= The ?123 digital signal obtains a reproduced image signal or a 125 output.乜 salt, from the output terminal [patent literature] special open 2001 - 2 〇 " 〇 2 bulletin = above, recorded, although the _ regenerative Wei Wei pain than the regenerative signal Tongle _ 1G7 equalization processing, also Can = a certain degree of characteristic differences, but for the larger special, ... ten further reduced errors but not enough. In particular, in recent years, j is a variety of magnetic tapes or magnetic heads, and the difference in characteristics is larger, = required performance Manga stabilization, adaptability to users, and cost reduction, equalizer equalization ^ Temple is more adaptable and can be adjusted with high precision.卞亚和-6- 1270059 [Summary of the Invention] The purpose of the line is to provide a data processing method which can soften and accurately carry out the regenerative decoration and equalization processing, thereby (4) reducing the reproduction error. Reproduction of the reproduced digital material, means for converting the analog signal of the reproducing means to a digital signal; and 'for the 15-bit equalization means from the class-like level converting means 10, the number two;:: , the number of the equalization processing number 2 _ financial number of the health _ vibrating key operation of the static wave i line;, from the variable filtering 11 means of the digital signal group delay variable group delay filter And, the digital signal from the variable group L / the wave is the hand 4 and the variable FM wave means. In the present invention in which the compensation is performed differently, the recording of the magnetic tape or the like is performed to obtain an analog signal, and in the case of such a ratio, the digital reproduction processing is performed, and the analog reproduction signal is converted into a number by the equalizer. Cheng Xu (4) Touch 2 rounds U./(1) Balanced hand magic implementation equalization processing. Number == means of equalization (the amplitude of the digitally-numbered bit signal (the means for the solution, the means for the wave device; the filter for adjusting the group of the digital signal late); and 4 cattle The relative group delays the chopping process of the digital processing and the target special pair 21 over these filter hands is different (or error) 20 1270059, 仏 filter means, and any filter means can be adjusted In this way, it is possible to perform a highly accurate equalization process in accordance with the characteristics of the recording medium such as a magnetic tape or the difference in characteristics of the reproducing head, etc. 5 10 15 In one embodiment, the variable filtering means has: a variable band pass filter means for adjusting a digital signal of the digital signal from the analog/digital conversion means; *, The variable high-pass variable band-pass filter means for adjusting the amplitude of the digital signal of the filter, so that the predetermined device in the digital signal passes through to adjust the amplitude characteristic, and then passes through the variable high-pass filter The high frequency in the number is preferentially adjusted to adjust the amplitude characteristic. Further, in one embodiment of the present invention, an interpolator in which the converted digital signal is interpolated and then sampled to obtain a materializing means is provided. = 抆 抆 进行 进行 进行 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内 内The FiR wave means means that the number L of the L number is connected to each other in series / η: the first stage or the middle of the plurality of latches are latched: the input coefficient and the digital bit are multiplied and output. Multiple @@目连接, the addition and addition of the two coefficient units, through the change of the characteristic of the hopper, by changing the coefficient === 20 1270059 tap coefficient The digital signal characteristics can be varied variously. The form of the change is arbitrary, for example, the variable filter means, the variable group delay filter means, and all the coefficients of the variable HR filter means are changed, and only the variable filter means is used. Coefficient change, only variable filtering Means and coefficient variation of the variable 5 group delay means. In the present invention, although the method of changing the coefficients of the various filter means is arbitrary, one of the methods is that the user writes the coefficient data value in the register, and the coefficient data is read from the register. [Embodiment] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings. Fig. 1 is a block diagram showing a configuration of a data reproducing device according to the present embodiment. The digital data recorded in a magnetic tape or the like is reproduced, amplified by an amplifier, and output to the analog filter 10. The analog filter 10 is an anti-aliasing filter and has an analog signal of fb/2 (fb ··bit rate) or more. The frequency components are truncated and output to the A/D 12. The filtering characteristics of the analog filter 10 are shown in the second figure. In the figure, the abscissa is the frequency, the ordinate is the amplitude, and fs is the sampling frequency. The A/D 12 converts the analog signal from the analog filter 10 into a digital 20 signal and outputs it to the interpolator 14. Specifically, A/D 12 samples the analog signal corresponding to the clock from the PLL not shown in the figure, and quantizes the digitization with one sample of multiple bits. The interpolator 14 determines the digital signal from the A/D 12 based on the poor point of the symbol point between the samples and the symbol point between the samples. -9 士 A/Dl2 'because the corresponding PLL A is engraved into the 仃 sampling (non-synchronized sample) ’ in the non-synchronous

: Insert the information of the symbol point. Interpolation: It is necessary to re-take the data after interpolation in the interpolator 14 to form a filter for the A/D. Water is free. The interpolator 丨4, which is basically shown by the FIR in the third figure, shows the configuration of the interpolator 14 in the interpolated crying; the filter characteristics of 'the lock in the fourth figure (9 in the figure) There are a plurality of 10 (a plurality of 10) coefficient units 14b-丨2 14^^ connected in parallel (the composition is 10 in the figure. Each latch 14a is only at -, 10 ' And the adder 14c and each of the coefficient units 14b-1 to Hb is called to hold the digital signal and rotate in the sample period. After multiplication, the output is added to the adder 14cc to adjust the coefficient and the input digital signal by 1 to Hb-10. The sum-adder 14c is output, and each of the coefficient units Hb is set to a digital equalizer 15 of the subsequent stage, for example, a coefficient set to ～10, which is set in advance in groups. The coefficient unit 14b-1, 2: the coefficient is 'Mb-3', the coefficient gas) 1 the coefficient unit 14b-4: the coefficient 2^ the coefficient unit 14b-5: the coefficient 2〇 coefficient unit 14b-6: the coefficient coefficient is 14b —7 : coefficient 1 coefficient 14b — 8 : coefficient two, coefficient state Mb —9, 1〇·· coefficient ~ and optional

Wait. The coefficient group is preferably in the prior Q group (for example, 32 groups '10^ 1270059). That is to say, one of the multiple groups is selected according to the position to be interpolated and added. The supplementary position should be interpolated, That is, the timing of the interpolation addition is adjusted by the timing control circuit composed of the timing error detector 34, the loop filter 36, and the NCO (Number Control Oscillator). 5 The digital signal interpolated and resampled by the interpolator 14 is The digital equalizer 15 is supplied to the digital equalizer 15. The digital equalizer 15 controls the amplitude of the digital signal and the group delay to match the digital signal with the desired target characteristic. In the present embodiment, the digital signal is equalized. The digital equalizer 15, specifically 10, comprises: a fixed FIR filter 16; a variable bandpass filter (variable BPF) 18; a variable high pass filter (variable HPF) 20; an all-pass filter 22; an automatic gain controller (AGC) 24; a variable FIR filter 26; and an adaptive controller 28. The fixed HR filter 16 provides a high frequency component for the digital signal from the interpolator 14. The deterioration of the high-frequency component is compensated for. In other words, since the high-frequency components of the original signal are deteriorated in the filters of the reproducing head 2, the analog filter 10, and the interpolator 14, only the high-frequency component is subjected to a predetermined value (fixed). The fifth graph shows the filter characteristics of the fixed HR filter 16. Among the illustrated characteristics, the coefficients of the respective coefficient units 20 of the fixed FIR filter 16 are set to be near fb/2. The frequency gain is increased. The digital signal compensated by the fixed FIR filter 16 is then supplied to the variable BPF 18. The variable BPF 18 is a filter that passes only the specified frequency components, and is capable of This frequency component is appropriately variably adjusted by filtering. Such a variable BPF 18, in a combination of a plurality of latches and coefficient units, also by multiplying the coefficients in the coefficient unit The configuration is variable. The filter characteristics of the variable BPF 18 are shown in the sixth figure, and the configuration of the variable BPF 18 is shown in the seventh figure. The variable Bm8 includes: (4) multiple (four in the figure) Latch (10); connected in parallel Two of the coefficient units 18b - 丨 彳 ( (Tita 丄 仙 1 2 2, and 'Adder 18c. The output of the 2m waver 16 is supplied to the latch 18a, and is also supplied to the coefficient. In the device 18b-i, the coefficient device (10)-- is multiplied with the variable coefficient Kb, and the bit 彳5唬 is outputted to the adder 18c after being chopped κ.... The stored 18a, keeping the digital signal '' will be Multiple latches cry to ^ Addition is 18 (: medium. In addition, the signal 2 is supplied to the signal 2, and the coefficient Γ is multiplied by the variable coefficient by half. in. The adder 18c outputs the signals of -1 . 18u 9 ^ 』 and HPF20 t after the signals are entered into the rod. By appropriately adjusting the variable system (4) shown in the figure of the coefficient unit 18b, it is possible to change the filter characteristics of the BPF as shown in the sixth figure. The arrow table in the figure changes: Ϊ; = Γ. It is used to characterize the ferrite of the variable ΒΡΠδ. The number Kb' can be appropriately supplied by the externally supplied adjustment signal, and is a filter through which the high frequency component passes, and the HPF 20, , , / / field is adapted to the appropriately variable adjustment filter. Such a variable transformer performs the combination of the latch 5 and the coefficient of the multiplier, and is also configured by making the coefficient of the system 去 I different. The 遽 and wave characteristics of the variable HPF2 ’ are shown in the eighth diagram. The setting in the ninth diagram -12-1270059 shows the configuration of the variable HPF 20. The variable addition also includes: a plurality of (two in the figure) latches 2〇a; a plurality of (two in the figure) coefficient units 20b-1, 20b-2; and an adder 2 〇c to form. The digital signal from the variable BpFi8 is supplied to the latch 20a and the coefficient unit. The coefficient unit 2Gb-L·Nujianjian variable arithmetic multiplication is calculated and output to the adder 20c. #is crying? ~^ A latched state 20a, after the digital signal is held, the input = adder 2〇c. In addition, the digital b tiger via the two latches 2〇a is supplied to the coefficient unit 2〇b-2, and the coefficient unit 2-3 pairs the input ίο 15 20 digital signal ^variation Kh for multiplication calculation Go to adder 1. Addition H 2〇e, after the signals are added and calculated, the output is in the all-pass filter 22. By appropriately adjusting the variable coefficient Kh of the coefficient unit 2% - 丨, 2, the filter characteristics of the HPF can be changed as in the first figure. The arrows in the figure indicate the change in ji ==. For changing the filter characteristics of variable HPF2Q, =

Kh 'is the same as the variable coefficient Kb in the variable BPF 18, and is appropriately set by the externally supplied adjustment signal. The cut is to pass the above-mentioned fixed FIR filter 16, variable Bppie to HPF2G' The reduction of the rape continues. . . . can be ... another - aspect 'all-pass filter 22, the main logarithm: the letter into the adjustment. The tenth figure shows the full money machine 22 & skin cry two tenth - figure t The configuration of the all-pass filter and the wave filter 22 = all pass ^ 2 'includes: the subtractor 22a; the delays 22b, 22d; the coefficient unit 22^ is one, ^1·22e HPF20 is subtracted into 22a. The subtractor 22a , the input digital signal and the difference from the delay -13 - l27 〇〇 59 I sign ^ is counted #, and output to the coefficient unit melon = towel. The coefficient of the melon, the difference signal and the variable coefficient a multiplication After the counting, the wheel is sent to the adder 22e and the delay 22d. The delay 〃^22 delays the tiger by only 1 sample and supplies it to the subtractor 22a, adding "the signal from the delayed state 22b and the signal from the coefficient 22c. Signal two or two, after the output. The subtractor melon, the coefficient unit 22C, and the delay ^1 IR filter are used to form the TM data filter by the coefficient unit melon, the delay unit, and the addition 1", and the group delay is controlled by adjusting the coefficient A of the coefficient unit. By setting the coefficient A of the coefficient unit 22c to a negative number, the group delay amount of the low-frequency component of the 10 3 large | A number , is changed by the value of the increase/decrease coefficient a as shown by the arrow in the figure. The coefficient A' of the all-pass filter 22 is also appropriately set by adjusting the signal from the outside, and the group delay of the k L-passer digitization number can obtain a slightly flat characteristic. 15 The digital signal compensated by the fixed FIR filter 16, the variable BPF 18, the variable HPF 20, and the all-pass m 22 is then supplied to the AGC 24 and then supplied to the variable FIR filter 26. . The AGC 24' adjusts the amplitude of the input digital signal supplied to the variable nR filter % to some extent, thereby preventing the amplitude of the variable nR chopper % 20 !!! 3 Therefore, for the variable FIR filter 26 The output is monitored with a gain error detection of 30, and the gain of the input digital signal is feedback adjusted by returning the output of the loop filter 32 to the AGC 24. The interchangeable FIR filter is 26, and is an FIR filter for matching the characteristics of the input digital signal with the target characteristic, and the coefficient of the coefficient is variable. The -14- 1270059 is a variable tilt, which is set by light adjustment. Self-adapted ^ b 28 adjustment signal to carry out the standard characteristics) and the round entry two: coffee according to the specified material ^ = 麟 ΐ ΐ, corresponding to the difference 5 10 15 twelfth figure = ^ Γ! number increase (four) whole. Characteristics, #β The drama "Characteristics of 旒" means: regeneration of $ 2 (10) characteristics of the ship's hiring machine 10, interpolation of the Dansheng head 2 wave device 1 Μ temple property, variable Β characteristics, _ = 2 〇 = ship filter Filter 22 tongues® 1 ΑΑ Α 4 inch, and whole god. 丌心τ重重(4)性' * The original σ and R filters, waver 26 and adaptive controller 28 are obtained by 敕= , 糸, so that the difference 5 of the two characteristics shown in the figure is eliminated. ° Positive, state touch twelve shows the variable filter 26 and the adaptive control cry 28 = Cheng. Variable m_6, including: The latches of the four series connected to each other: four are connected to each other. In addition, the adaptive controller 28 includes a determiner 2 such as a subtractor Mb and an adaptive algorithm processor 28c. The coefficient unit 26b of the variable nR filter % is divided into four at 4 to 2, and the adjusted coefficient is multiplied with the wheeled digital signal, respectively, and the result is output to the adder 26c. The adder 26c adds the addition. The calculation result is outputted to the Viterbi decoder 42 of the next stage, and is also output to the determiner 28 & and the subtractor 28b of the adaptive controller 28. The coefficients of 26b-1 to 26b-4 are initially set to a certain value by an external adjustment signal, and then adjusted to an appropriate value by the adjustment signal of the adaptive algorithm processor 28c. The determiner 28a, the adder The output of 26c is compared with the threshold value, and -15- 20 1270059 determines which one of the specified digit values is the round-out digit value from the adder 26c (when the digit value of the shape is G, -b +1, The digit value from the addition 26c is compared with the threshold value to determine which of the values. For example, when the output digit value from the adder 26c is 0.8, the determiner 28a determines that it is +1 and outputs it to the subtracter. The subtracter 28b' subtracts the output from the adder 26c and the decision result from the determiner 28a to calculate the difference. The difference or difference is the difference between the characteristic of the input digital signal and the target characteristic. The winning difference is rounded out to the adaptive algorithm processor 28c. 10 The adaptive algorithm processor 28c performs an LMS (least mean square ([caffeine==)) algorithm, ie, minimizes the square of the difference (error signal) Can == each of the coefficients 26 of the wave H 26 The coefficient of 丨~施―4 is always changed. The adaptive algorithm processor tears up with the circuit, but it can also be used for software processing. Lai, can _ lose the number of letters (four) = /, 111 waver 26 The variable coefficient, although adjusted by the above-mentioned adaptive H-type through the feedback feedback control, can also be adjusted by adding or subtracting the coefficient directly through the register °. A suitable implementation method is that The external adjustment signal (20 from the register is applied to each coefficient of the variable FIR waver 26), and the initial value of h is set, and then the coefficient value is adjusted by an appropriate algorithm.

5 Tigers show the characteristics (the coffee feature) to achieve high-speed equilibrium convergence.

The digit number which is finally equalized by the FIR filter 26 is supplied to the Viterbi decoder 42. The Viterbi decoder 42 detects the digital signal by the _-16-1237059 and outputs it to the signal processing circuit 44. The signal processing circuit 44 obtains a reproduced image signal or a reproduced sound signal based on the digit number obtained by the Viterbi decoder 42 and outputs it to the display device. 5 10 15 20, the output of the variable FIR filter 26 is supplied to the timing error detector 34 in addition to the gain error detection cry 30, and is supplied to the NC port via the loop filter σ 36 (Number c〇) The ntr〇1 〇sdllat〇r) 38 generates a control signal corresponding to the time error and supplies it to the interpolator 14 for two weeks. The interpolator M, based on the control signal from the NC 038, interpolates as a group of corresponding control signals. "" As described above, the digital signal of the data reproduction and interpolator 本 of the present embodiment is subjected to the equalization process by the digital equalizer 15, and the 'by each of the digital equalizers 15' is A filter (a filter capable of adjusting its characteristics = a characteristic difference of a recorded tape or the like, or a difference in the reproduction head, a characteristic difference of the class 10, etc., is softly and accurately compensated.

In the present embodiment, the HPF2 is variable with a fixed FIR chopper 16 m. , all-pass chopper 22, and variable. heart; constitute digital balance E丨5, with fixed money ιίί ΒΡ ^, variable HPF2 〇 mainly adjust the amplitude of the digital signal, and use eight ^ two i to adjust the digital signal The group delay is supplied to the variable S, and thus, the digits supplied to the variable TM filter 26 are close to the target touch, and (4) at -17-

When the target characteristic is equalized, it is possible to suppress the divergence of the divergence due to the extreme deviation of the input number two characteristics. The change of the FIR filter 26 after the group delay control is performed by the second companion group delay controller 26 alone. The group delay control capability is acceptable = Since the all-pass filter a is implemented in the present embodiment: the system is late control, the variable FIR comparator 26 only controls the all-pass reduction and the high-precision control. In other words, variable compensation 8, variable HPF2〇, all-pass can be realized; the wave filter ^ (is compensated by the digital signal of the digital signal, and = the machine is clear, (10) ^ ^ to 15 or more For example, the definition is changed to HPF20 = more realistic: the method is described in the seventh embodiment, and the invention is not in the embodiment, but is configured to make the variable BPF 18, the all-pass filter state 22, the versatile, the Lu, the shi, and the (four) 20 The coefficient (tap coefficient) is adjusted by external adjustment: two signals), but it is expected that the remainder (four) coefficient two, for example, writes the appropriate residual data value in the register, and the number of resources is: from this register to each coefficient In the device. In addition, in the present embodiment, it is not necessary to adjust all the coefficients Ί8-1270059 of the variable Βρ HPF20, the all-pass wave $22, and the variable FIR filter % to change their characteristics, for example, The variable BPF 18 is maintained as a default feature and only the coefficients of the variable HPF 20 and all-pass filter 22 are adjusted. Of course, it is also possible to adjust only the coefficients of the variable BPF 18, or only the coefficients of the variable HPF 20, or only the coefficients of the all-pass filter 22. In the present embodiment, the high-frequency component degraded by the processing of each component of the reproducing head 2, the analog filter 10, and the interpolator 14 is compensated by the fixed FIR filter 16, but may be compensated. The fixed FIR filter 16 is omitted corresponding to the characteristics of the reproducing head 2 and the like. The data reproduction device of the present embodiment can be combined with a DVC (Digital Imaging 10) or HDD (Hard Disk Drive), a CD drive, or a DVD drive reproduction device, and can reproduce digital data such as PR4 as an analog signal. It is suitable for use in any device that digitally reproduces such a reproduction signal.

-19- 1270059 [Simple description of the drawings] The first figure is an overall configuration diagram of the embodiment. The second figure is a characteristic explanatory diagram of the analog filter in the first figure. The third figure is a characteristic explanatory diagram of the interpolator in the first figure. 5 The fourth figure is a configuration diagram of the interpolator in the first figure. The fifth figure is a characteristic explanatory diagram of the HR filter in the first figure. Fig. 6 is a characteristic explanatory diagram of the variable BPF in the first figure. Fig. 7 is a configuration diagram of the variable BPF in the first figure. The eighth figure is a characteristic explanatory diagram of the variable HPF in the first figure. 10 The ninth diagram is a configuration diagram of the variable HPF in the first figure. The tenth diagram is a characteristic explanatory diagram of the all-pass filter in the first figure. The eleventh diagram is a configuration diagram of the all-pass filter in the first figure. Fig. 12 is an explanatory view showing the difference between the digital signal characteristics and the target PR4 characteristics of the respective filters. 15 Figure 13 is a block diagram of the variable FIR filter and the adaptive controller in the first figure. Figure 14 is a configuration diagram of a conventional device. 1270059 [Description of main component symbols] 2 Regeneration head 10 analog filter 12 A/D 14 interpolator

14a Latch 14b-1~14b-10 Coefficient Unit 14c Adder 15 Digital Equalizer 16 Fixed F1R Filter 18 Variable BPF 10 15

18a latch 18b-1, 18b-2 coefficient unit 18c adder 20 variable HPF 20a latch 20b-1, 20b-2 coefficient unit 20c adder 22 all-pass filter 22a subtractor 22b, 22d delay unit 22c Coefficient device

22e adder 24 AGC 26 variable HR filter 26a latch 26b-1~26b-4 coefficienter 26c adder 28 adaptation controller 28a determiner 28b subtractor 28c adaptation algorithm processor 30 gain error detector 32, 36 Loop filter

34 timing error detector 38NCO 42 Viterbi decoder 44 signal processing circuit 20

Claims (1)

1270059 I. Patent application scope: 1. A data reproduction device, comprising: a regeneration means for reproducing digital data; converting an analog signal from the regeneration means into a digital ratio/digital conversion means, and H negative 5 10 15 20, a digital equalization means for performing equalization processing on a desired target characteristic on the digital signal from the analog/digital conversion means, /, the data equalization means having ···from the analog/digital a variable filter period of the digital signal adjustment of the conversion means; (4) a variable group delay filter means for adjusting the digital signal from the variable chopper means; and performing the pair from the variable group The variable π R filter means for compensating for the difference in the digital signal target characteristics of the delay filter means, wherein the variable filter means has: adjusting the amplitude of the digital signal of the === segment to perform the digitization of the line __ The feature is that the π 玍 玍 玍 ,, the filter means has a digital letter from the analog/digital conversion means a fixed band HR filter for adjusting the high frequency band -22-1270059; a variable band pass filter means for adjusting a predetermined band amplitude of the digital signal from the fixed HR filter means; A variable high-pass filter that re-adjusts the high-band 5 amplitude of the digital signal of the variable bandpass filter. 4. The data reproducing apparatus according to claim 1, further comprising: interpolating and re-sampling the digital signal from the analog/digital conversion means, and obtaining the obtained digital signal Output to the interpolator in the 10-digit equalization means. The data reproducing device according to claim 1, wherein the variable filter means, the variable group delay filter means, and the variable F1R filter means have: 15 for inputting a digital signal a plurality of latches connected in series with each other; a plurality of coefficient units connected to a front stage or an intermediate or a rear stage of the plurality of latches, and multiplying a coefficient by an input digital signal; and An addition unit 20 that adds the outputs from the plurality of coefficient units changes the characteristics of the filter by making the coefficients variable. The data reproducing device according to claim 5, further comprising: a register for variably setting the coefficient, -23-1270059. The filter characteristic is characterized by writing a coefficient data value in the register. The variable FIR filter means according to the above-mentioned claim 1 is: Which variable of the HR filter of the variable system that is multiplied by the variable coefficient and multiplied by the variable coefficient is the difference of the value of the 3-value digit value (4) the coefficient of the output of the determiner is The variable characteristic of the variable coefficient device according to the data reproducing device described in the first item, wherein the variable chopper means and the variable group β bit signal and the target characteristic are The difference is delayed by filtering, means, and the digital data set forth in item 1 of the patent application scope to be levied is the digital data pre-committed by PR4. -twenty four,